Modern interactive media (e.g., wikis, social tagging systems, multi-player games, augmented realities) allow capturing rich streams of data about student actions while learning. Analyzing this multi-dimensional data can provide novel insights on learning and assessment for both individuals and collaborative teams. How can this type of diagnostic, formative assessment resolve current limitations of summative high-stakes tests. This evolution of assessment is crucial for moving curricula and instruction towards 21st century understandings and performances. And that is the working premise of Chris Dede’s presentation.
Professor at Harvard University, but that’s just the tipping point of his accompishments. Emerging virtual world and virtual performance assessment is part of Dede’s approach to the shifting world of educational assessment. Dede begins with noting that tests often penalize students in many ways because real knowledge is not necessarily measured. Education is held accountable to meeting measures, whatever they may be, but current summative tests undercut achievement and motivation.
From How People Learn to Knowing What Students Know provides a core idea: assessment is a detective kind of situation where you try to find what students really know. We move from cognition and observations to interpretation of the data/observations. Newer models, however, are more sophisticated. The problem in paper-and-pencil testing is that it doesn’t measure higher cognitive skills. The old triangle is replaced by collaborative, mediated, scaffolded, and data-generated event logs.
Take a look at this video.
Privacy is disappearing, for better or worse, because we are surrounded by tools that capture what we are doing. However, these tools open up tools for educational assessment for emerging technologies because they tell us how our students are learning. Social networking and virtual worlds provide rich enviroments with event logs. What evidences of authentic assessment of student work can we utilize. In addition to those listed, we have active log file data. Watch this video from River City. The point Dede wants to make is that they have the opposite of what most teachers have in the classroom. They have the opposite problem: enormous data–too much intellectually digestible data–to help them learn what students are really doing via scientific inquiry. While it is a powerful learning environment, it also generates a powerful assessment environment in a cognitive model within a 21st century skill.
Dede admits that River City was really not equipped to assess the data that the virtual world created. Their newest venture tackles the assessment piece that is richer than a physical performance assessment or an item test by putting them into a rich virtual world.
Take a look at this video on virtual assessment. Dede notes that is is an early version that they have already superceded in some ways. For example, in a virtual environment, you explore why Celts are dying by examining all aspects of the unhealthy environment, thus causing death. You use the NSES Model of Inquiry to make sense of what they investigate, and then the assessment tracks how the students learn in the virtual environment, which is better than paper-and-pencil environments/tests.
If we were really good at formative assessments weekly with students, we would not need summative assessments, but we just do not have a great snapshot of what students really know. So, you model a virtual ecosystem in a real ecosystem, hoping that people will transfer learning from virtual systems to real systems.
Watch this video on this interface and the data they hope to generate from the virtual ecosystem. Science and causality are important effects of learning in a virtual environment like the Pond Ecosystem Module. Learning trajectories exist for the learning modules to track data. Less rich but valuable forms of chat files by media in education are important for summative assessment and we need to begin to mine it.
As Dede addresses future initiatives, he mentions that he feels that breakthroughs are coming in the next 4-5 years, but we should not wait for that. We should start now and make our own breakthrough. It’s not just about tests and grades, but what happens in the real world. A test should not be a narrow measure but rich among multiple dimensions. We need not to teach to the test but look at new forms of assessments that can replace them.
In Q&A session, Dede suggests that we need to begin to hold discussions with parents in our learning communities to take a look at how we assess what students know. Dede mentions that includes A students who have mastered the grading game, but sometimes experience difficulty in the real world, where paper-and-pencil high-stakes tests are not the measure of authentic life tasks.
When asked why such a push exists in science, he notes that STEM is where the money is, so science becomes where they begin, but not necessarily where this type of learning ends, because there is broad applicability in history and other disciplines. Educators who think like Sherlock Holmes or Columbo have the same ability to transfer various curricular areas into virtual educational environments.